To quantify how strict are gun controls in different states, we are using:
| variable | type | complete | mean | missing | n | p0 | p100 | p25 | p50 | p75 | sd |
|---|---|---|---|---|---|---|---|---|---|---|---|
| application_rate | numeric | 50 | 0.1 | 0 | 50 | 0.0 | 1.0 | 0.1 | 0.1 | 0.1 | 0.1 |
| approval_rate | numeric | 50 | 0.0 | 0 | 50 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
| crude_rate | numeric | 50 | 13.2 | 0 | 50 | 3.6 | 23.9 | 10.3 | 12.5 | 17.3 | 4.9 |
| death_rate | numeric | 50 | 13.0 | 0 | 50 | 3.4 | 23.0 | 9.7 | 12.8 | 17.4 | 4.9 |
| death_rate_rank | integer | 50 | 25.5 | 0 | 50 | 1.0 | 50.0 | 13.2 | 25.5 | 37.8 | 14.6 |
| deaths | numeric | 50 | 771.0 | 0 | 50 | 49.0 | 3353.0 | 206.0 | 564.0 | 1048.2 | 745.9 |
| law_strength | integer | 50 | 25.4 | 0 | 50 | 1.0 | 50.0 | 13.2 | 25.5 | 37.0 | 14.5 |
| number_of_background_checks | numeric | 50 | 498647.7 | 0 | 50 | 12742.0 | 4641480.0 | 129223.0 | 299839.0 | 534990.5 | 720686.3 |
| number_of_federal_firearms_licensees | numeric | 50 | 2718.7 | 0 | 50 | 246.0 | 10920.0 | 1228.0 | 2247.5 | 3566.0 | 2111.0 |
| number_of_residents_in_millions | numeric | 50 | 6.5 | 0 | 50 | 0.6 | 39.5 | 1.8 | 4.6 | 7.3 | 7.3 |
| population | numeric | 50 | 6448926.9 | 0 | 50 | 585501.0 | 39250017.0 | 1850105.5 | 4559320.0 | 7198767.8 | 7271769.1 |
| state_code | numeric | 50 | 29.3 | 0 | 50 | 1.0 | 56.0 | 17.2 | 29.5 | 41.8 | 15.8 |
| year | numeric | 50 | 2016.0 | 0 | 50 | 2016.0 | 2016.0 | 2016.0 | 2016.0 | 2016.0 | 0.0 |
| variable | type | complete | empty | max | min | missing | n | n_unique | ordered |
|---|---|---|---|---|---|---|---|---|---|
| death_cause | character | 50 | 0 | 7 | 7 | 0 | 50 | 1 | NA |
| grade_2017 | factor | 50 | NA | NA | NA | 0 | 50 | 5 | 0 |
| state | character | 50 | 0 | 14 | 4 | 0 | 50 | 50 | NA |
| state_abb.x | character | 50 | 0 | 2 | 2 | 0 | 50 | 50 | NA |
| state_abb.y | character | 50 | 0 | 2 | 2 | 0 | 50 | 50 | NA |
As we can see, the dataset has no missing data. All the information, including population statistics, background check / applications statistics, approved licensees statistics, are generalized for different states (except for D.C.).
The first plot show the relationship between the percentage of people get approved for gun licenses, mortality rate and the law strength in each state.
As we can see in this plot, in different states, the percentage of people get approved for gun licenses is positively connected with law strength. This is intuitive because it’s easier to get approved for guns in states with less strict gun laws. Also, with less strict gun laws, states generally have higher mortality rate. An regression model is built to further explain this relationship in model panel.
The second plot shows the relationship application percentage (proportion of people applying for the license in total population), mortality rate and the law strength.
As we can see in the plot,